A claimed proof of the Riemann hypothesis runs 250,000 pages.

If it holds, it settles the Clay Institute’s $1 million prime-number problem after 167 years.

Anthropic’s Levent Alpöge says he built it with Claude, and its Lean translation runs 17 million lines.

Cornell’s Steven Strogatz says no one has independently verified the claimed proof.

A computer check of the Lean code comes first, and Scott Aaronson says full understanding may take years.

How each outlet framed it

drawn from 100+ reports worldwide

New Scientist leans favorable
credits Anthropic's Levent Alpöge and a Claude-assisted proof, and frames Hughes's caveat that the result is still unverified
Yahoo leans favorable
quotes Alpöge's 'Claude flew, and danced' account and Strogatz's note that the computer check has sometimes erred

Sources: New Scientist, Yahoo